No abstract

No abstract

The appearance of flowing fluid has been evaluated in several clinical situations using a flow phantom, computer simulation, and clinical magnetic resonance (MR) images. Unsaturated protons just entering the imaging volume can emit a strong signal relative to the partially saturated adjacent tissue ("flow-related enhancement"). Slow laminar flow in veins can be distinguished on the basis of a stronger second echo due to rephasing effects ("even echo rephasing"). Synchronization of the cardiac cycle and the MR pulsing sequence produces increased signal in sections acquired during diastole ("diastolic pseudogating"). Intraluminal signal is shown to decrease as velocity is increased ("high velocity signal loss"). Onset of turbulence causes further loss of signal. Direction of flow oblique to the imaging plane can be predicted on the basis of decreased upstream and increased downstream signal.

Magnetic resonance imaging (MR) (0.35 T) and computed tomography (CT) were compared in 400 consecutive patients with suspected disease of the brain and cervical spinal cord. Of 325 positive diagnoses, MR detected abnormality while CT was normal in 93; MR was more specific in 68; MR and CT gave equivalent information in 129; CT was more specific in 32; and CT was positive while MR was normal in 3. MR was superior to CT in detection of multiple sclerosis, subcortical arteriosclerotic encephalopathy, posterior fossa infarcts and tumors, small extra-axial fluid collections, and cervical syringomyelia. CT was preferable in evaluation of meningiomas and separation of tumor from edema. Using available sequence parameters, MR missed 3 1-cm calcified meningiomas which were clearly seen on CT. CT takes less time and may be preferable in patients with acute trauma as well as very young or elderly individuals. Thus the two studies should be considered complementary.

The appearance of rapidly flowing blood on imaging (MRI) was evaluated using flow phantoms and dye infusion experiments. Laminar flow can be maintained at high velocities in small-diameter vessels. Under such condftions, flow-related enhancement may be observed several slices into a multislice imaging volume. Decreasing crosssectional area of the unsaturated protons in the midstream is noted on slices further removed from the entry surface.

These radiologic changes may help distinguish ELSTs from other tumors of the temporal bone and posterior fossa.

No abstract

Knowledge of the exact dose and rate at which an antitumor agent is delivered to its target site is postulated to be crucial to proper patient management. It is now possible to obtain such information using non‐invasive 19F‐NMRS (nuclear magnetic resonance spectroscopy) following the administration of 5‐fluorouracil (5‐FU). We have performed such studies in 103 patients with breast, colorectal and other tumors. Measurable 19F signals were detected in 99 of these patients (92.5%). Estimation of the tumoral t1/2 of 5‐FU in these patients revealed that 51 of them (51.5%) exhibited a tumoral t1/2 greater than 20 min, a value we had characterized as indicating drug trapping in the tumor. Of these patients, 46 who received regimen bolus 5‐FU 600 mg/m2 with leucovorin for their treatment have been evaluated. In these patients, the association between trapping and response remains very high (p < .000001). None of the non‐trappers responded to chemotherapy, whereas 70% of the evaluable trappers responded. Details are presented here on the methodology of NMRS data acquisition and on their pharmacokinetic analysis. The potential mechanisms underlying the trapping effect appear to be predicated primarily on transport processes. Suggestions are presented on how such pharmacokinetic imaging studies may extend both our understanding of the mechanism of action of 5‐FU, and how they could be used to optimize patient treatment. © 1998 John Wiley & Sons, Ltd.